Optical Satellite Signal Processing and Enhancement Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Optical Satellite Signal Processing and Enhancement Shen-En Qian SPIE PRESS Bellingham, Washington USA Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Library of Congress Cataloging-in-Publication Data Qian, Shen-En. Optical satellite signal processing and enhancement / Shen-En Qian. pages cm Includes bibliographical references and index. ISBN 978-0-8194-9328-6 1. Image processing. 2. Imaging systems–Image quality. 3. Signal processing. 4. Remote-sensing images. 5. Optical images. I. Title. TA1637.Q48 2013 629.43'7–dc23 2013006792 Published by SPIE—The International Society for Optical Engineering P.O. Box 10 Bellingham, Washington 98227-0010 USA Phone: +1 360 676 3290 Fax: +1 360 647 1445 Email: [email protected] Web: http://spie.org Copyright © 2013 Society of Photo-Optical Instrumentation Engineers (SPIE) All rights reserved. No part of this publication may be reproduced or distributed in any form or by any means without written permission of the publisher. Thecontentofthisbookreflectstheworkandthoughtoftheauthor(s).Everyefforthas beenmadetopublishreliableandaccurateinformationherein,butthepublisherisnot responsible for the validity of the information or for any outcomes resulting from reliance thereon. Printed in the United States of America. First printing Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Contents Preface xv List of Terms and Acronyms xix 1 Spaceborne Optical Sensors 1 1.1 Introduction 1 1.2 Optical Satellite Sensors and Their Types 2 1.3 Panchromatic Sensors 2 1.4 Multispectral Sensors 5 1.4.1 Landsat MSS, TM, and ETMþ 5 1.4.2 SPOT’s HRV, HRVIR, and HRG 9 1.4.3 Other multispectral sensors 10 1.5 Hyperspectral Sensors 11 1.5.1 What is a hyperspectral sensor? 11 1.5.2 Operating principle of a hyperspectral sensor 12 1.5.3 Types of hyperspectral sensors 13 1.5.3.1 Dispersing-element-based sensors 14 1.5.3.2 Optical-filter-based sensors 14 1.5.3.3 Electronically tunable-filter-based sensors 15 1.5.4 Hyperspectral sensor operating modes 16 1.5.4.1 Whisk-broom mode 16 1.5.4.2 Push-broom mode 17 1.5.5 Spaceborne hyperspectral sensors 18 1.5.5.1 Ultraviolet and Visible Imagers and Spectrographic Imagers system 18 1.5.5.2 Hyperion 18 1.5.5.3 Compact High-Resolution Imaging Spectrometer 20 1.5.5.4 Medium-Resolution Imaging Spectrometer 20 1.5.5.5 Compact Reconnaissance Imaging Spectrometer for Mars 21 1.5.5.6 Moon Mineralogy Mapper 22 1.5.5.7 Advanced Responsive Tactically Effective Military Imaging Spectrometer 23 1.5.5.8 Environmental Mapping and Analysis 23 v Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms vi Contents 1.6 Imaging Fourier Transform Spectrometer Sensor 23 1.6.1 Description 23 1.6.2 Types of FTS sensors, and operational concept 24 1.6.3 Spaceborne IFTS 25 1.6.3.1 Infrared Atmospheric Sounding Interferometer 26 1.6.3.2 Tropospheric Emission Spectrometer 27 1.6.3.3 Cross-track Infrared Sounder 27 1.6.3.4 AtmosphericChemistryExperiment–Fourier TransformSpectrometer 28 1.6.3.5 Fourier Transform Hyperspectral Imager 28 1.6.3.6 ASTRO FTS 29 1.6.3.7 Geosynchronous Imaging Fourier Transform Spectrometer 29 1.7 Lidar Sensor 30 1.7.1 Definition and description 30 1.7.2 Lidar In-space Technology Experiment 32 1.7.3 Shuttle Laser Altimeter 33 1.7.4 Mars Orbiter Laser Altimeter 34 1.7.5 Geoscience Laser Altimeter System 34 1.7.6 Cloud-Aerosol Lidar with Orthogonal Polarization 34 1.7.7 Atmospheric Laser Doppler Lidar Instrument 35 1.7.8 Mercury Laser Altimeter 36 1.7.9 Lunar Orbiter Laser Altimeter 37 1.7.10 Next-generation, high-resolution swath-mapping lidar 38 References 39 2 Satellite Data Generation and Product Levels 43 2.1 Space Data and Information System 43 2.2 EOS Data and Information System 43 2.2.1 Spacecraft command-and-control center 44 2.2.2 Data capture and Level-0 processing 44 2.2.3 Product generation 44 2.2.4 Data archive, management, and distribution 45 2.2.5 Locating and accessing data products of interest 45 2.3 EOS Data Product Levels 45 2.4 Planetary Data System and Product 46 2.4.1 Standard data products 46 2.4.2 Engineering and other ancillary data products 47 2.4.3 Dataset documentation 48 2.5 Planetary Data Product Levels 49 2.6 Example of EOS Data Product Levels 49 2.6.1 Level-0 data products 50 2.6.2 Level-1 data products 50 2.6.3 Level 2 and higher data products 50 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Contents vii 2.7 Example of Planetary Data Product Levels 51 2.7.1 Level 1: Raw data 52 2.7.2 Level 2: Raman datasets 52 2.7.3 Level 3: Calibrated unidentified Raman spectra 52 2.7.4 Level 5: Carbon/mineralogy results 53 2.7.5 Level 6: Ancillary data 53 2.7.6 Level 7: Correlative data 54 2.7.7 Level 8: User description 54 References 54 3 Satellite Data and Image Quality Metrics 55 3.1 Needs for Quality Metrics 55 3.2 Full-Reference Metrics 57 3.2.1 Conventional full-reference metrics 57 3.2.1.1 Mean-square error (MSE) 57 3.2.1.2 Relative-mean-square error (ReMSE) 58 3.2.1.3 Signal-to-noise ratio (SNR) 58 3.2.1.4 Peak signal-to-noise ratio (PSNR) 58 3.2.1.5 Maximum absolute difference (MAD) 58 3.2.1.6 Percentage maximum absolute difference (PMAD) 58 3.2.1.7 Mean absolute error (MAE) 59 3.2.1.8 Correlation coefficient (CC) 59 3.2.1.9 Mean-square spectral error (MSSE) 59 3.2.1.10 Spectral correlation (SC) 60 3.2.1.11 Spectral angle (SA) 60 3.2.1.12 Maximum spectral information divergence (MSID) 61 3.2.1.13 ERGASformultispectralimageafterpan-sharpening 61 3.2.2 Perceived-visual-quality-based full-reference metrics 61 3.2.2.1 Universal image quality index 61 3.2.2.2 Multispectral image quality index 62 3.2.2.3 Quality index for multi- or hyperspectral images 64 3.2.2.4 Structural similarity index 65 3.2.2.5 Visual information fidelity 67 3.3 Reduced-Reference Metrics 68 3.3.1 Four RR metrics for spatial-resolution-enhanced images 70 3.3.2 RRmetricusingwavelet-domainnatural-imagestatisticmodel 72 3.4 No-Reference Metrics 74 3.4.1 NR metric for compressed images using JPEG 75 3.4.2 NR metric for pan-sharpened multispectral image 76 3.4.2.1 Spectral distortion index 77 3.4.2.2 Spatial distortion index 77 3.4.2.3 Jointly spectral and spatial quality index 78 References 78 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms viii Contents 4 Satellite Data Compression 81 4.1 Lossless and Near-Lossless Data Compression 81 4.1.1 Lossless compression 82 4.1.2 Near-lossless compression 83 4.2 Vector Quantization Data Compression of Hyperspectral Imagery 85 4.2.1 Review of fast VQ compression algorithms 85 4.2.2 Near-lossless VQ compression techniques 89 4.2.2.1 Successive approximation multi-stage vector quantization 89 4.2.2.2 Hierarchicalself-organizingclustervectorquantization 90 4.3 Onboard Data Compression of Multispectral Images 91 4.3.1 1D differential pulse code modulation 91 4.3.2 Discrete-cosine-transform-based compression 91 4.3.3 Wavelet-based compression 93 4.3.4 Selective compression 94 4.4 Lossless Compression of Ultraspectral Sounder Data 95 4.4.1 Comparison of wavelet-transform-based and predictor-based methods 96 4.4.1.1 Wavelet-transform-based methods 96 4.4.1.2 Predictor-based methods 97 4.4.1.3 Comparison results 98 4.4.2 Lossless compression using precomputed vector quantization 99 4.4.2.1 Linear prediction 99 4.4.2.2 Grouping based on bitlength 100 4.4.2.3 Vector quantization with precomputed codebooks 100 4.4.2.4 Optimal bit allocation 100 4.4.2.5 Entropy coding 101 4.4.3 Lossless compression using the prediction-based lower triangle transform 101 4.4.3.1 Prediction-based lower triangle transform 102 4.4.3.2 PLT lossless compression algorithm 103 4.4.3.3 Results of PLT lossless compression 104 4.5 CCSDS Data Compression Standards for Spacecraft Data 106 4.5.1 Three space-data compression standards 106 4.5.2 Lossless data compression standard 107 4.5.3 Image-data compression standard 110 4.5.4 Lossless multispectral/hyperspectral compression standard 114 References 118 5 Satellite Data Formatting and Packetization 125 5.1 Formatting Satellite Data Using CCSDS Space Data Link Protocol 125 5.2 Telemetry System Concept 127 5.2.1 Packetization layer 128 5.2.2 Transfer frame layer 128 5.2.3 Channel coding layer 129 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Contents ix 5.3 Space Packet Concept 129 5.4 Space Packet Structures 130 5.4.1 Packet primary header 131 5.4.1.1 Packet version number 131 5.4.1.2 Packet identification field 132 5.4.1.3 Packet sequence control field 132 5.4.1.4 Packet data length 133 5.4.2 Packet datafield 133 5.4.2.1 Packet secondary header 133 5.4.2.2 User datafield 134 5.5 Telemetry Transfer Frame 134 5.5.1 Transfer frame primary header 135 5.5.1.1 Master channel identifier 136 5.5.1.2 Virtual channel identifier 136 5.5.1.3 Operational control field flag 136 5.5.1.4 Master channel frame count 136 5.5.1.5 Virtual channel frame count 137 5.5.1.6 Transfer frame datafield status 137 5.5.2 Transfer frame secondary header 138 5.5.3 Transfer frame datafield 139 5.5.4 Operational control field 140 5.5.5 Frame error control field 141 References 142 6 Channel Coding 145 6.1 Telemetry System Layers and Channel Coding 145 6.2 Channel Coding Improving Space Data Link Performance 147 6.2.1 Channel coding performance measures 147 6.2.2 Shannon limit on channel coding performance 148 6.3 Reed–Solomon Codes 150 6.3.1 Definition 150 6.3.2 RS encoder 152 6.3.3 Interleaving of the RS symbols 154 6.3.4 Decoding of RS codes 155 6.3.5 Performance of RS codes 156 6.4 Convolutional Codes 157 6.4.1 Encoder for CCSDS (7, 1/2) convolutional code 157 6.4.2 Encoder for CCSDS punctured convolutional code 159 6.4.3 Soft maximum-likelihood decoding of convolutional codes 160 6.4.4 Performance of (7, 1/2) code and punctured convolutional codes 161 6.5 Concatenation of Reed–Solomon and Convolutional Codes 163 6.6 Turbo Codes 166 6.6.1 Definition 166 6.6.2 Turbo encoder and decoder 167 6.6.3 Comparing turbo codes to traditional concatenation codes 169 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms x Contents 6.7 Low-Density Parity-Check Codes 171 6.7.1 Introduction 171 6.7.2 CCSDS-recommended LDPC codes 172 6.7.2.1 Base (8176, 7156) LDPC code 173 6.7.2.2 Shortened (8160, 7136) LDPC code 174 6.7.3 Performance of LDPC code 175 References 176 7 Calibration of Optical Sensors 179 7.1 Importance of Calibration 179 7.2 Absolute and Relative Radiometric Calibration 181 7.3 Satellite Optical Sensor Modeling 184 7.4 On-Ground Calibration prior to Launch 186 7.4.1 Review 186 7.4.2 Landsat instrument laboratory calibration 188 7.4.3 AVIRIS laboratory calibration 188 7.5 Onboard Calibration Postlaunch 190 7.6 Vicarious Calibration 193 7.7 Conversion to At-Sensor Radiance and Top-of-Atmosphere Reflectance 195 7.7.1 Conversion to at-sensor radiance 195 7.7.2 Conversion to top-of-atmosphere reflectance 196 7.7.3 Conversion to at-sensor brightness temperature 197 References 198 8 Keystone and Smile Measurement and Correction 205 8.1 Keystone and Smile in Imaging Spectrometers 205 8.1.1 Spectral distortion: smile 205 8.1.2 Spatial distortion: keystone 208 8.1.3 How keystone and smile affect pixel shape and location 210 8.2 Method of Measuring Smile Using Atmospheric-Absorption Feature Matching 212 8.3 Smile Measurements of Five Hyperspectral Imagers 216 8.3.1 Testing AVIRIS sensor smile 216 8.3.2 Smile measurement of the SFSI sensor 219 8.3.3 Smile measurement of the CASI sensor 222 8.3.4 Smile measurement of the CHRIS sensor 225 8.3.5 Smile measurement of Hyperion 225 8.4 Measuring Keystone Using Interband Correlation of Spatial Features 231 8.5 Measuring Keystone of Hyperspectral Imagers 233 8.5.1 Test of keystone of AVIRIS sensor 233 8.5.2 Measuring keystone of the Aurora sensor 235 8.5.3 Measuring keystone of the CASI sensor 236 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms Contents xi 8.5.4 Measuring keystone of the SFSI sensor 237 8.5.5 Measuring keystone of the Hyperion sensor 237 8.5.6 Summary of keystone measurement results 238 8.6 Effects of Keystone on Spectral Similarity Measures 239 References 241 9 Multisensor Image Fusion 243 9.1 Image Fusion Definition 243 9.2 Three Categories of Image Fusion Algorithms 246 9.3 Conventional Image Fusion Methods 247 9.3.1 IHS fusion 247 9.3.2 PCA fusion 251 9.3.3 Arithmetic combination fusion 252 9.3.4 Wavelet transform fusion 254 9.4 Comparison of Typical Image Fusion Techniques 257 9.4.1 Brief description of nine fusion techniques 257 9.4.2 Summary of evaluation results 259 9.5 Image Fusion Using Complex Ridgelet Transform 261 9.5.1 Purpose 261 9.5.2 Radon transform 262 9.5.3 Ridgelet transform 262 9.5.4 Operation of iterative back-projection 264 9.5.5 Image fusion based on the complex ridgelet transform 264 9.5.6 Image fusion experimental results 267 9.6 Fusion of Optical and Radar Images 274 9.6.1 Fusion of multispectral and SAR images using intensity modulation 275 9.6.2 SARandopticalimagefusionbasedonwavelettransform 276 9.6.3 SAR and optical image fusion based on local variance and mean 276 9.6.4 Fusion of RADARSAT-1 and SPOT images 277 References 279 10 Enhancing the Spatial Resolution of a Satellite by Exploiting the Sensor’s Keystone Distortion 289 10.1 Enhancing Satellite Sensor Performance Using a Signal Processing Approach 289 10.2 Exploiting the Keystone of a Satellite Sensor to Enhance Spatial Resolution 291 10.3 Using Keystone to Increase the Spatial Resolution of a Single-Band Image 294 10.3.1 Fusion of subpixel-shifted images 294 10.3.2 Method 1: Separate band images extracted based on KS-induced subpixel shift 296 Downloaded From: http://ebooks.spiedigitallibrary.org/ on 10/12/2014 Terms of Use: http://spiedl.org/terms